Human Fanconi Anemia Complementation Group A Protein Stimulates the 5’ Flap Endonuclease Activity of FEN1

In eukaryotic cells, Flap endonuclease 1 (FEN1) is a major structure-specific endonuclease that processes 5’ flapped structures during maturation of lagging strand DNA synthesis, long patch base excision repair, and rescue of stalled replication forks. Here we report that fanconi anemia complementation group A protein (FANCA), a protein that recognizes 5’ flap structures and is involved in DNA repair and maintenance of replication forks, constantly stimulates FEN1-mediated incision of both DNA and RNA flaps. Kinetic analyses indicate that FANCA stimulates FEN1 by increasing the turnover rate of FEN1 and altering its substrate affinity. More importantly, six pathogenic FANCA mutants are significantly less efficient than the wild-type at stimulating FEN1 endonuclease activity, implicating that regulation of FEN1 by FANCA contributes to the maintenance of genomic stability.


Introduction
During DNA replication, various structured DNA intermediates are effectively processed to avoid deleterious consequences such as genome instability and cancer. Replication on the lagging strands is discontinuous and initiated by DNA Pol α, which synthesizes an RNA primer approximately 12-nt long that is further extended with approximately 20-nt of DNA. This RNA-DNA hybrid is believed to be made at low fidelity and subjected to displacement by the high-fidelity DNA Pol δ. The strand displacement activity of Pol δ produces a 5' single-stranded flap structure that contains the RNA primer and some of the initiator DNA. The 5' flap structure is then recognized and incised by the structure-specific flap endonuclease FEN1 to ensure the integrity of DNA replication [1][2][3][4][5][6][7][8][9][10]. Additionally, FEN1 has 5' to 3' exonuclease activity and gap-dependent endonuclease activity that are important during maturation of Okazaki fragments and the rescue of stalled replication forks [6,[11][12][13][14][15][16]. FEN1 is also involved in long-patch base excision repair, nucleotide excision repair, nonhomologous end-joining, and resolution of di-and tri-nucleotide repeat secondary structures [7,[17][18][19]. Defects in FEN1 cause accumulation of mutations, genomic instability, cancer predisposition and chronic inflammation [20][21][22].
FANCA has been found to exist in a large nuclear protein complex that contains the Fanconi anemia core complex and the Bloom syndrome BLM complex [36,43]. It is localized to chromatin in a replication-dependent manner [44][45][46]. Xenopus egg extracts immune-depletion study shows that FANCA is directly involved in maintenance of replication forks [47,48]. Most recently, FANCA has been found to regulate MUS81-EME1 activity in a damage-dependent manner and FANCA has intrinsic affinity to nucleic acids with particularly high affinity to single-stranded RNA and DNA structures with a 5' flap [46,49]. Because both FANCA and FEN1 localize to replication forks and practically share the same substrate specificity, i.e. 5' flap structures and single-stranded RNA [44][45][46], we hypothesize that FANCA may affect FEN1 activity by competing or collaborating with each other on the same structures to regulate removal of RNA primers and 5' flap structures during DNA replication and repair. In support of this hypothesis, deficiency in FEN1 causes partially similar phenotypes as FANCA, i.e. inflammation and cancers [22,[50][51][52].
In this study, we found that human FANCA indeed stimulates FEN1 endonuclease activity. This novel regulation of FEN1 by FANCA is impaired in pathogenic FANCA mutants thus making the novel interaction physiologically relevant to Fanconi anemia.

Materials and Methods
Expression and Purification of Proteins cDNAs for human FANCA and FEN1 were obtained by PCR amplification from a universal cDNA pool (BioChain Institute, Inc.). The full-length open reading frames were confirmed by sequencing and found to exactly match NCBI Reference Sequence NM_000135 and NM_004111 respectively. Overexpression and purification of hexahistidine-tagged FANCA was achieved in insect High Five cells using the Bacto-Bac expression system (Invitrogen) as previous described [16,46]. Truncation mutants of FANCA were produced through a PCR-based method [53]. Point mutations were produced through a site-directed Mutagenesis Kit (Agilent). Expression of FANCA and its mutants was confirmed by Western blot analysis using FANCA Antibody (Santa Cruz Biotech.). Monoclonal antibody against the His6 tag (GenScript, Piscataway, NJ) was also used to confirm expression and subsequent purification. Protein concentration was determined using the Coomassie protein assay reagent (Pierce). The purified proteins were stored at -80 °C in aliquots. Purified FEN1 was prepared as described previously [13,54] and confirmed by western blot using a FEN1 antibody (Epitomics).

Preparation of Substrates
Oligonucleotides that were used to create the 15-nt 5' flap substrates were adopted from a design by Fisher et al with the same sequences ( Figure S1) [55]. RNA/DNA hybrid oligos were chemically synthesized by Integrated DNA Technologies, Inc. with the flap as RNA. All DNA oligos were purified by 10% denaturing PAGE gel. The 5' ends in the flap structures were labeled by 32 P ( Figure S1). Annealing was carried out in a water bath within 5 h by slowly cooling from 70 °C to 20 °C.

Endonuclease Assay
The endonuclease assay was performed as previously described [1]. 2 nM of 5′ 32 P-labeled 5' flap substrates were incubated with purified proteins as indicated amount in a 10 μl reaction with the buffer containing 30 mM HEPES PH 7.5, 1 mM dithiothreitol, 3 mM MgCl 2 , 5% glycerol, 100 ng/mL bovine serum albumin and 100 mM KCl at 37 °C for 15 minutes. The reaction was stopped by adding 10 μl 2x sequencing dye (10 mM EDTA, 0.2% SDS, 0.03% Xylene cyanol and Bromophenol blue). Reaction products were separated on a 10% or 15% denaturing polyacrylamide gel. The incision products were visualized by autoradiography and quantified by using NIH ImageJ software. The incision rate was calculated by dividing the intensity of product band by the total substrate band of each reaction.

Determination of Kinetic Parameters
To measure kinetic parameters, kinetic analyses were repeated three times using increasing amounts (described under the figure) 15nt both DNA and RNA 5' flap substrates. Kinetic parameters were obtained based on the Michaelis- where v is the reaction rate and [S] is the concentration of substrates. K m and V max were gained by plotting v against [S] using Origin software through nonlinear curve fit.

FEN1 incises 5' RNA flap differently from the DNA counterpart
In order to test whether FANCA interacts with FEN1, we overexpressed and purified full-length human FEN1 protein ( Figure S1A). Considering the length of the in vivo substrate of FEN1, we designed 15-nt 5' DNA and RNA flaps ( Figure S1B). Initial incubation of the purified protein with the flap structures showed that FEN1 incised both the DNA and RNA flap structures ( Figure 1A). Intriguingly, FEN1 incised the DNA flap differently from the RNA counterpart. FEN1 had two major incision sites on the 15-nt DNA flap with one right at the junction site ( Figure 1A, DNA panel, arrow 2) and the other at -1 base inside the junction site ( Figure 1A, DNA panel, arrow 1). However, FEN1 cut the RNA flap only at the -1 position inside the junction ( Figure 1A, RNA panel). Furthermore, FEN1 incised the RNA flap significantly more efficiently than the DNA flap ( Figure 1A, compare lanes 1-7 with 8-14; Figure 1B).

FANCA stimulates the 5' flap endonuclease activity of FEN1
Since FANCA and FEN1 share the same substrate specificity, i.e. 5' flap structures and single-stranded RNA [46], we hypothesize that FANCA physically and functionally interacts with FEN1. Physical interaction between FANCA and FEN1 has never been reported previously. In order to examine whether FANCA interacts with FEN1 in cells, we prepared whole-cell extracts of the FANCA-null and FLAG-FANCAcomplemented cells (FANCA -/-and +/+ respectively in Figure  2A). Next, we performed a co-immunoprecipitation assay of the extracts using a mouse anti-FLAG antibody and detected FANCA and FEN1 using a goat anti-FANCA and a rabbit anti-FEN1 antibodies respectively (Figure 2A, top panel). As shown in Figure 2A, When FANCA was pulled down by the anti-FLAG antibody, FEN1 followed, indicating that FANCA interacts with FEN1 in cells. To confirm the physical interaction, we performed the co-immunoprecipitation assay using a FEN1 antibody (Figure 2A, bottom panel). Again, FANCA was steadily detected in the pull-down lysate. These results indicated that FANCA and FEN1 interact with each other in human cells.
To test whether FANCA functionally affects the catalytic activity of FEN1, we purified human wild-type FANCA to near homogeneity ( Figure S1A). Next, we titrated the purified FANCA in a flap endonuclease assay using suboptimal amounts of FEN1 ( Figure 2B, 0.2 nM for DNA flap and 0.1 nM for RNA flap respectively). To rule out the possibility of the stabilizing effect of proteins on FEN1, we diluted purified FEN1 and FANCA proteins in a buffer with 1 μg/μl BSA. Incision of the 15-nt DNA and RNA flaps by FEN1 alone is ~11% and ~4% of the total substrate respectively. However, addition of increasing amount of FANCA greatly enhanced the flap endonuclease activity of FEN1 by up to 8-fold for RNA flap and 4.5-fold for DNA flap within the titration range ( Figure 2B). These data establish that FANCA physically interacts with FEN1 and functionally stimulates the flap endonuclease activity of FEN1 in a concentration-dependent manner.

FANCA increases the enzyme efficiency of FEN1
To determine how FANCA affects the flap endonuclease activity of FEN1, we performed a steady-state analysis of FEN1 by titration of the DNA and RNA flap substrates in the presence or absence of FANCA ( Figure S2). The obtained incision rate (v) and the substrate concentration [S] were fit into the Michaelis-Menten equation v = V max [S] / (K m +[S]) in a nonlinear manner ( Figure 3). As shown in Figure 3, FANCA resulted iñ 4-5 fold increase in FEN1 enzyme turnover rate on the DNA flap (k cat. 0.125 vs. 0.026). However, FANCA reduced the DNA flap substrate affinity of FEN1 (K m 13.5 vs. 33.9). Overall, FANCA increased the enzyme efficiency of FEN1 by 2-fold (kcat/K m 1.9 pM -1 s -1 vs. 3.7 pM -1 s -1 ). A similar result was obtained with the RNA flap substrate (Figure 3), i.e., FANCA increased the turnover as well as K m of FEN1 with a 1.7-fold overall increase in enzyme efficiency. This result is distinct from that of PCNA which decreased the K m of FEN1 [25], as well as that of WRN or RFC, which increased the V max but did not alter substrate binding [28,57].

Both N-and C-terminals of FANCA are required for the stimulation of FEN1 activity
We previously showed that FANCA has a nucleic acid binding domain at the C terminal and this domain confers the preferential binding of FANCA to ssRNA, ssDNA, and 5' flaps [58]. To test whether the nucleic acid binding domain of FANCA affects the stimulation of FEN1 by FANCA, we used two truncation mutants of FANCA, Q772X and C772-1455 for FEN1 assay. Q772X is a Fanconi anemia disease-causing Cterminal truncation mutant. C772-1455 is the complementing C-terminal fragment of Q772X (Fig. 4A). Using 10 nM of protein that is sufficient for the WT protein to exert its stimulation, we found that both mutants showed drastic reduction in stimulating FEN1 endonuclease activity ( Figure 4B and 4C, last two lanes in each panel). These results indicate that both the N-terminal and the nucleic acids binding C-terminal of FANCA are indispensable for FEN1 stimulation and that the DNA binding domain by itself is insufficient to regulate FEN1 activity.

Pathogenic FANCA mutants are significantly less efficient in stimulating FEN1 activity
To evaluate whether stimulation of FEN1 by FANCA is physiologically relevant to Fanconi anemia, we created 5 more FANCA point mutations and purified them to near homogeneity (Fig. 4A). D598N, R951W, R1055W, R1117G, and F1263Δ are selected from FANCA mutations that cause Fanconi anemia [59,60]. D598N, R951W, R1055W, and R1117G are pathogenic missense point mutations. F1263Δ is a one residual deletion mutant representing one of the most prevalent pathogenic point mutations. Using 10 nM of protein, we found that all of the FANCA disease-causing mutants have defects in stimulating the endonuclease activity of FEN1 ( Figure 4B and 4C). These results clearly demonstrate that mutations in FANCA significantly affect its ability to stimulate FEN1 and the interaction between FANCA and FEN1 is relevant to the etiology of Fanconi anemia.

Discussion
In this study, we aimed to unveil the function of the preferential 5' flap and RNA binding activity of FANCA [46] and found a novel physical and functional interaction between FEN1 and FANCA. We demonstrated that FEN1 incises a RNA flap more efficiently than its DNA counterpart and FANCA further stimulates this incision catalyzed by FEN1. More importantly, all six pathogenic FANCA mutant proteins we tested were defective in this interaction to different degrees, indicating that the interaction between FEN1 and FANCA  Intriguingly, before we study how FANCA may affect FEN1 activity, we observed that FEN1 alone, different from a previous study reporting that FEN1 does not cleave the 5' RNA flap structure [61], incises the 5' RNA flap more efficiently than its DNA counterpart (Figure 1). This is obviously beneficial to its function in RNA primer removal during maturation of Okazaki fragment. Additionally, the incision pattern for the RNA flap is different from the DNA flap ( Figure 1). It showed that FEN1 only cleaves the RNA flap substrate at the -1 position inside the junction ( Figure 1A, RNA panel) which is actually a DNA base pair. However, there were two cleavage sites on the DNA flap substrate: one at the junction site and the other at -1 base inside the junction site ( Figure 1A, DNA panel). This is different from a previous study showed that mammalian FEN1 cuts DNA flaps only at the -1 position inside the junction site [61]. Our data supports that the nature of the 5' flap affects the incision sites and efficiency of FEN1 initially proposed by Bambara's group [62].
Besides the physical interaction we showed in Figure 2 and the shared substrate specificity of FEN1 and FANCA for 5' flap and RNA, our confocal microscopy result showed that FANCA perfectly colocalizes with replication forks in unstressed human cells (Qian et al, unpublished data), indicating that FANCA is associated with the normal replication machinery where FEN1 is known to be found for regular maintenance of replication forks. Additionally, it is estimated that more than 10 6 replication-stalling DNA lesions per cell per day form in humans. Because the number of Okazaki fragments per cell cycle is about 20-50 X 10 6 in humans, one replication fork stalling event may associate with about 20-50 Okazaki fragments theoretically [16,29,63]. This estimate suggests that FEN1 is likely to encounter and interact with FANCA that is recruited to maintain stability of replication forks [44][45][46]. Our in vitro result with the 15-nt 5' RNA and DNA flaps demonstrated that FANCA is likely to be involved in the removal of RNA/DNA primers by facilitating FEN1 action during maturation of Okazaki fragments.
FEN1 participates in the long-patch base excision repair of non-bulky DNA lesions (oxidation, methylation, base loss, etc) by interacting with Pol β, APE1, Lig 1, PCNA and Neil1 for efficient removal of the damaged bases and 5' flaps [17,18,64,65]. FANCA was also shown to be involved in base excision repair through stabilizing the glycosylase Neil1 [66]. Our result indicated that FANCA additionally interacts with FEN1 to facilitate removal of 5' DNA flap catalyzed by FEN1. Our findings suggest that the Fanconi anemia pathway may directly regulate the excision repair of DNA lesions caused by oxidative stress explaining the oxidative stress sensitive phenotype of Fanconi anemia cells [67].
Because of the preferential binding of FANCA to 5' flap structures [46], we hypothesized that FANCA may facilitate loading FEN1 to the 5' flap substrate and therefore increase the substrate affinity of FEN1. However, FANCA increased both the turnover rate and K m of the endonuclease activity of FEN1 ( Figure 3). Based on these results, we speculate that FANCA may regulate the endonuclease activity of FEN1 through two possible mechanisms: (i) Direct protein-protein interaction between FANCA and FEN1 that changes FEN1 conformation and increases its endonuclease turnover and (ii) competition for the 5' flap substrate between FANCA and FEN1 that reduces substrate affinity of FEN1. It is conceivable that reduced substrate affinity of FEN1 in the presence of FANCA may also cause faster release of the incision product and therefore help FEN1 to turnover. Overall, these possible mechanisms result in about two-fold increase in FEN1 enzyme efficiency. This possible mechanism is supported by the data that both non-nucleic-acids-binding N-and nucleic-acidsbinding C-terminals of FANCA are required for FEN1 stimulation (Figure 4). It remains to be determined how FANCA exactly interacts with FEN1, but it is distinct from WRN, RFC, and PCNA in interacting with FEN1. Both WRN and RFC increased the turnover of FEN1 without affecting the substrate affinity, on the other hand, PCNA increased the substrate affinity of FEN1 without changing the turnover rate [25,28,57].
The next interesting question remaining to be answered is whether FANCA affects other activities of FEN1. Like FANCA, the gap-dependent endonuclease as well as the exonuclease of FEN1 is important for rescue of stalled replication forks [7,47,48]. It has been shown that defects in the gap-dependent endonuclease and exonuclease of FEN1 cause chronic inflammation and cancers [22]. Coincidently, deficiency in FANCA also results in inflammation and cancers [50][51][52]. Based on the physical and functional interactions between FANCA and FEN1, it is conceivable that FANCA may regulate the gap-dependent endonuclease as well as exonuclease activities of FEN1 and therefore contribute to suppression of inflammatory responses and maintenance of genomic stability.